We had officer elections at the queer sorority today. I am the person who takes minutes and in charge of community service activities. (I asked, community service includes raising money for charity. :D) The elections themselves had an interesting format:
Each candidates for a given position makes a speech.
All members of the sorority besides the contenders for the can ask individual candidates questions.
Then candidates are removed from the room and everyone inside the room talks about what they think about the candidates. People put forth arguments and try to reach a state in which no one feels like their position will change.
The candidates are brought back in and everyone closes their eyes and votes. The president counts up the votes and then includes their own.
This system worked well here. I think it would work well other places given that there is not time pressure and people cannot repeat the same argument if people find it annoying unless someone consents. I was surprised how smoothly it worked. Everyone was civil and no one tried to either (1) repeat their position to people who have said that this argument would not change their mind or (2) belittle the arguments of others. This system allows for people to reflect and some to consider better arguments for each candidate than were stored in their minds. I am still unsure about the process of deciding on a political candidate based on majority vote (using preference voting) from members of the constituency. I am working on alternatives.
However, allowing people to argue until they do not anticipate their opinion changing and then voting, if done with rules to prevent people from being annoying, is an improvement over the alternative. That is, it has less expected regret than a system in which votes are cast immediately after the candidates have explained their positions. To see this, assume we have a situation in which everyone besides the candidates waits until after the election and then argues over who is the best person for the job until everyone anticipates their opinion being consistent. Clearly a group who had done this before the election experiences strictly less regret than one who has not. This is not only relevant to this thought experiment. It seems reasonable that this argument session models a formalized and sped up process of debate that will happen between members throughout the course of the administration. The debate also allows people to form their best guesses at what will happen if each candidate takes power. Doing some kind of observed trial run before the election would be better. However, it is difficult and time-consuming to design a procedure for all of the, here, 17 officer positions. Even if they were designed people would argue that results are not evidence for or against competency at the given job. The session effectively immunizes everyone against experiencing regret over a large swath of arguments made by people who have detailed knowledge of how the community works. This is definitely added to my CEV toolbox.